Skip to main content

CompletionModel

rig_cat::model

Trait CompletionModel 

Source 
pub trait CompletionModel {
    // Required methods
    fn complete(
        &self,
        request: CompletionRequest,
    ) -> Io<Error, CompletionResponse>;
    fn stream(&self, request: CompletionRequest) -> Stream<Error, StreamChunk>;
}
Expand description

The core LLM abstraction: send a request, get a response.

All provider-specific logic is behind this trait. Returns Io<Error, _> for composable effect handling.

Required Methods§

Source

fn complete(&self, request: CompletionRequest) -> Io<Error, CompletionResponse>

Send a completion request and get a full response.

Source

fn stream(&self, request: CompletionRequest) -> Stream<Error, StreamChunk>

Send a completion request and get a streaming response.

Implementors§